The present invention generally relates to computer memory systems, and more specifically, to repurposing of spare locations in a memory system to store internally-generated data.
Memory subsystems may make use of multiple types of error checking, correction and reporting mechanisms to improve recoverability, availability, and serviceability (RAS) and data integrity. In higher end systems, such as mainframes, it is typical for data to be checked, corrected and reported upon in many places within write and read paths. The write and read paths can include checking, correction and reporting from a memory controller to a physical memory interface through chip packaging, board routing across a board into a memory device, and return paths. As memory technology has evolved, the amount of checking, correction, and reporting included in memory subsystems has expanded to enable higher speed and higher density technologies. Higher density technologies can be more prone to certain issues, such as charge loss.